Add more AI-generated code review examples

Background

AI-generated code review is one of the most important parts of the v2 handbook. The current content explains the principles well, but more concrete examples would help developers apply the workflow.

Scope

Add or improve examples covering:

• Large AI diff review
• Splitting AI changes into logical commits
• Detecting unrelated file changes
• Reviewing generated tests
• Writing PR descriptions for AI-assisted changes
• Deciding when to ask a human reviewer to stop and split the change

Possible target files:

• 05-ai-native-development/ai-change-review-example.md
• 05-ai-native-development/ai-generated-code-review.md
• 05-ai-native-development/ai-commit-splitting.md

Done When

• Examples are realistic and copyable.
• Risk checks are concrete, with file or diff-level thinking.
• The English version is updated when needed.
• python3 scripts/check-docs.py passes.